WindBorne Systems
WindBorne Systems designs, manufactures, and operates the Atlas constellation of long-duration weather balloons and the WeatherMesh AI forecasting model, selling proprietary atmospheric data and forecasts to NOAA, the U.S. Air Force, U.S. Navy, energy traders, and utilities.
- Company typePrivate
- Founded2019
- HeadquartersRedwood Shores, United States
- Headcount51–100
- GTM typeB2B
- OfferingSoftware
What WindBorne Systems does
WindBorne Systems is a private atmospheric-intelligence company that designs, manufactures, and operates the Atlas constellation — a network of long-duration, autonomously navigated Global Sounding Balloons (GSBs) — and the WeatherMesh family of AI weather forecasting models. Founded in 2019 by Stanford Space Initiative alumni John Dean, Dr. Andrey Sushko, and Kai Marshland, the company is headquartered in Redwood City, California, with approximately 50 employees and global launch operations spanning roughly 15–16 sites across three continents. Each 1.2 kg solar-powered GSB flies for weeks (12+ days typical, with demonstrated 75+ day missions) producing around 50 controlled vertical atmospheric profiles per flight, generating in-situ data over oceans, polar regions, and the Southern Hemisphere where traditional radiosonde and satellite systems leave major gaps.
The company's core technology stack combines three proprietary layers: (1) the Atlas GSB constellation as the underlying observation source, (2) a proprietary AI Data Assimilation (AI DA) system that ingests 11 observation types independent of ECMWF/NOAA schedules, and (3) the WeatherMesh transformer-based forecast model, currently in its sixth generation (WM-6 Global at 0.25° / 25 km resolution plus a 3 km HighRes variant for CONUS and Europe). Adjacent products include MetaMesh (a learned multi-model blender of WeatherMesh, HRRR, GFS, IFS, and AIFS), WindBorne Guardian (an outage-prediction SaaS for utilities), and an Energy Trading forecast stack. Delivery is via REST API with JSON and Zarr responses, a Python SDK ("pip install windborne"), and a sales-led enterprise motion.
WindBorne monetizes through a mix of subscription / usage-based API access for commercial customers (energy traders, hedge funds, utilities), fixed-price U.S. government contracts (NOAA, U.S. Air Force, U.S. Navy, Defense Innovation Unit, DoD), and philanthropic / development-funded deployments (Gates Foundation for African agricultural forecasting). Total disclosed venture funding is approximately $25 million at a $85 million reported valuation, with no Series B announced as of mid-2026 despite active scaling toward a 10,000-balloon target constellation.
WindBorne Systems firmographics
Firmographics- Name
- WindBorne Systems
- Legal name
- WindBorne Systems Inc.
- Website
- https://windbornesystems.com
- Company type
- Private
- Founded year
- 2019
- Operating status
- Operating
- Headcount range
- 51–100 employees
- Short description
- WindBorne Systems designs, manufactures, and operates the Atlas constellation of long-duration weather balloons and the WeatherMesh AI forecasting model, selling proprietary atmospheric data and forecasts to NOAA, the U.S. Air Force, U.S. Navy, energy traders, and utilities.
- Ownership category
- akta.pro rank
WindBorne Systems industry classification
Industry- Product category
- AI Weather Intelligence / Atmospheric Data Platform
- NAICS
- Wind Electric Power Generation (221115), Geophysical Surveying and Mapping Services (54136)
- SIC
- Search, Detection, Navagation, Guidance, Aeronautical Sys (3812)
- akta.pro primary industry
- Wind Resource Assessment & Energy Yield Analytics (Met Masts, LiDAR, Modeling) (EUAMACAJ)
Keywords
Where WindBorne Systems is headquartered
LocationHeadquarters
- HQ city
- Redwood Shores
- HQ country
- United States
- HQ region
- North America
Offices2 records
Markets served
WindBorne Systems business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Personnel, Technology or R&D, Supply Chain, Infrastructure, Operations, Marketing or Sales
Revenue model
- Atmospheric Observation Data (Data-as-a-Service): Sale of raw and processed atmospheric data collected from the Atlas constellation to government agencies, defense, and commercial customers via REST API. Includes 1-year historical backtest data for some products. Customers include NOAA (since early 2023), U.S. Air Force, U.S. Navy, and commercial sectors.
- AI Weather Forecast API (WeatherMesh & MetaMesh): Subscription/usage-based access to AI forecast products (WeatherMesh-6 Global, WeatherMesh-6 HighRes, MetaMesh) delivered via REST API with JSON and Zarr responses, plus a Python SDK. Includes backtest data licensing, custom variable queries, and insights platform access. Targeted at energy traders, hedge funds, and utilities.
- Government Contracts: Fixed-price government awards and contracts with the U.S. Navy (six-month contract for subseasonal/tropical cyclone forecasting), U.S. Air Force (557th Weather Wing, 16th Weather Squadron, Air Force Lifecycle Management Center), NOAA, and Defense Innovation Unit (DIU) for AI forecasting R&D and operational model deployment.
- Industry-Specific Weather Solutions (e.g., Guardian for Utilities): Tailored SaaS-style solutions combining AI forecasts with vertical-specific data — e.g., WindBorne Guardian outage prediction platform for utilities. Free waitlist access available for the Guardian starter product; enterprise tiers with add-ons available.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Subscription | Annual | Energy Trading API access — WeatherMesh-6 Global, HighRes, and MetaMesh via REST/Python SDK |
| Hybrid | Annual | WindBorne Guardian for Utilities — starter free waitlist + enterprise add-ons |
Go-to-market motion4 records
Distribution channels6 records
Marketing channels10 records
WindBorne Systems product offering
Product offeringCore offering
WindBorne Systems designs, manufactures, and operates a constellation of long-duration Global Sounding Balloons (GSBs) called Atlas to collect in-situ atmospheric data over oceans, polar regions, and remote areas where traditional radiosondes do not reach. It then feeds that proprietary data into its own transformer-based AI forecasting model (WeatherMesh) and multi-model blender (MetaMesh), selling forecast and observation outputs to government agencies, defense customers, energy traders, and utilities through a REST API, Python SDK, and tailored vertical solutions.
Product overview
WindBorne Systems offers a planetary intelligence platform-plus-modules architecture that combines proprietary atmospheric data collection with AI-powered forecasting. The core platform consists of Atlas (the global balloon constellation powered by Global Sounding Balloons) for atmospheric observations and the WeatherMesh family of AI forecast models (including WeatherMesh-6 Global, WeatherMesh-6 3km/HighRes, and WeatherMesh-5c) for AI predictions. Built on top of this core, the platform extends with MetaMesh — a multi-model blend incorporating HRRR, GFS, IFS, AIFS, and WeatherMesh inputs — and three industry-specific solutions: a Weather Solutions layer for general industry applications, an Energy Trading forecast stack (combining WeatherMesh-6-Global, WeatherMesh-6 HighRes, and MetaMesh) for trading desks, and WindBorne Guardian for utility outage prediction. The platform is supported by the WindBorne Live data viewer for government partners and delivered to customers via the WeatherMesh Python SDK and a public REST API (with JSON and gridded Zarr response formats) backed by an interactive Insights platform.
Differentiator
Problem solved
Functional benefit
Brands
- Atlas: WindBorne's global atmospheric sensing constellation composed of long-duration Global Sounding Balloons (GSBs) - described as the world's largest balloon constellation, the foundation of the company's planetary nervous system.
- WeatherMesh
- MetaMesh
- WindBorne Guardian
- WindBorne Live
Products and services
- Atlas (Global Sounding Balloon Constellation) WindBorne's patented global balloon constellation of long-duration Global Sounding Balloons (GSBs) that collect in-situ atmospheric soundings over oceans, polar regions, and remote areas; produces approximately 50 controlled vertical profiles per GSB versus a single profile from a traditional weather balloon and operates with multi-week autonomous flight.
- WeatherMesh AI Forecast Model Family Transformer-based deep learning AI weather forecasting model family running on GPUs, trained on 50+ years of atmospheric data and continuously updated with Atlas observations; produces medium-range global forecasts up to 100,000x faster than physics-based NWP models.
- WeatherMesh-6 Global Latest-generation WeatherMesh operating at 0.25-degree (~25km) resolution, producing a 128-member ensemble in latent space with hourly forecasts out to 15 days and 163 total variables; up to 38% lower ensemble-mean RMSE than ECMWF IFS.
- WeatherMesh-6 3km (HighRes) High-resolution 3km variant of WeatherMesh-6 covering CONUS and Europe; assimilates GOES satellite imagery and MRMS radar composites; forecasts out to 72 hours, refreshing every 15 minutes, and outperforming NOAA HRRR at every forecast hour beyond hour 0.
- MetaMesh Deterministic multi-model blender that dynamically weights HRRR, GFS (and ensemble), IFS (and ensemble), AIFS (and ensemble), and WeatherMesh, with weights varying by lead time, variable, region, and season and recalibrating daily; provides hourly point forecasts out to 15 days optimized for 5,349 METAR stations.
- Energy Trading Forecast Stack Weather intelligence product built for energy traders, combining WeatherMesh-6 Global, WeatherMesh-6 HighRes, and MetaMesh into a single API-delivered stack with rolling updates, hourly timesteps out to 15 days, sub-second latency, and 1-year backtest data; delivered via REST API and Python SDK with an interactive Insights platform.
- WindBorne Guardian Disaster preparedness platform for climate-resilient utilities that combines AI weather forecasts with utility data (asset health, vegetation risk, historical outages) to predict weather-driven outages 0-14 days in advance, with hourly-refreshed neighborhood-scale predictions, crew staging optimization, critical-customer prioritization, and asset-specific failure risk identification.
- Weather Solutions Industry-tailored solutions layer that transforms WindBorne's proprietary balloon data and AI forecasts into actionable insights and automations for weather-impacted teams across energy trading, utilities, and emergency services.
- WindBorne Live Real-time data viewer that enables government partners, regulators, and customers to view the global GSB constellation, individual balloons, search soundings by geographic region, and compare GSB readings to NOAA and GFS model data; used by hundreds of meteorologists and partners including NORAD, NOAA, DoD, NCAR, the U.S. Air Force, and Scripps.
- WeatherMesh Python SDK & REST API Programmatic interface to WindBorne's WeatherMesh forecast models, MetaMesh blended forecasts, and Atlas observation data via a REST API with JSON and gridded Zarr response formats and a Python SDK (pip install windborne), with authentication via WB_CLIENT_ID and WB_API_KEY environment variables and sub-second latency.
Companies that use WindBorne Systems
Customer profileNamed customers9 records
Segments5 records
Ideal customer profiles4 records
WindBorne Systems technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration5 records
AI capability9 records
Feature7 records
WindBorne Systems partnerships and signals
Strategic signalPartnerships
13 partnerships are on record, tiered core — philanthropic/philanthropic-funder partner, flagship — featured in openai startup spotlight, core — federal defense partner, flagship — federal defense customer, core — institutional scientific partner, flagship — operational customer since early 2023, operational vendor, core — gpu/compute partner, core — academic research collaborator and core — airspace/governmental partner.
- Gates Foundationcore — philanthropic/philanthropic-funder partnerGates Foundation partnership referenced on WindBorne's homepage "Proven Partnerships" section; press coverage in Observer describes the partnership on weather forecast balloon AI, with related coverage of WindBorne balloons improving African forecast systems.
- OpenAIflagship — featured in openai startup spotlightWindBorne was featured in "OpenAI Presents | Startup Spotlight: WindBorne" (6.5.26) — a marketing/spotlight feature from OpenAI showcasing WindBorne's use of OpenAI technology.
- Defense Innovation Unit (DIU)core — federal defense partnerListed among WindBorne's ongoing government partnerships that the new Department of Defense awards build upon. Complements USAF/USN contracts and NOAA work.
- U.S. Navyflagship — federal defense customerIn July 2024, WindBorne began a six-month contract with the U.S. Navy to expand WeatherMesh into subseasonal forecasting and optimize the model for tropical cyclone prediction. The contract accelerates infrastructure tests to establish compute requirements for augmenting WeatherMesh to meet the Navy's operational needs.
- U.S. Air Force (557th Weather Wing, 16th Weather Squadron, Air Force Lifecycle Management Center)flagship — federal defense customerWindBorne has partnered with the U.S. Air Force to develop AI forecasting solutions that are already outperforming top AI weather models including Huawei's Pangu-Weather and Google's GraphCast. Includes extending WeatherMesh with an ensemble and high-resolution forecasting, demonstrating edge computing feasibility, and applying AI/deep learning to data assimilation.
- ECMWF (European Centre for Medium-Range Weather Forecasts)core — institutional scientific partnerECMWF logo displayed in WindBorne's "Proven Partnerships" section on the homepage; listed as one of the partners WindBorne is proud of, alongside government, private sector, and philanthropic organizations.
- NOAA (National Oceanic and Atmospheric Administration)flagship — operational customer since early 2023WindBorne data is distributed globally via WMO's Global Telecommunication System and assimilated into NOAA's Global Forecast System (GFS). NOAA studies have shown WindBorne data improves the accuracy of physics-based forecast models. WindBorne also lists NOAA as a logo on the "Proven Partnerships" section of its homepage.
- Better Stackoperational vendorPer WindBorne blog (7.22.25): "WindBorne Systems ensures global fleet reliability with Better Stack" — observability/monitoring infrastructure used to support WindBorne's global balloon constellation.
- DigitalOceanoperational vendorPer WindBorne blog (9.17.25): "Powering Next-Generation Weather Forecasting with DigitalOcean" — cloud infrastructure customer story.
- Zulipoperational vendorPer WindBorne blog (7.14.25): "Zulip x WindBorne: Case Study" — team collaboration/chat tooling adopted by WindBorne.
- NVIDIAcore — gpu/compute partnerWindBorne's WeatherMesh transformer-based AI models run on NVIDIA GPUs; WindBorne featured at NVIDIA CEO Keynote at CES 2025 (1.6.25). Coverage in Heatmap News notes "Windborne, Google DeepMind, NVIDIA" together in AI weather forecasting context.
- Scripps Institution of Oceanography (CW3E)core — academic research collaboratorCollaborative outreach and research event at Scripps Pier (6.30.25); WindBorne x CW3E Collaboration on validating sensors (5.29.25); Scripps regularly uses WindBorne Live tracking data.
- North American Aerospace Defense Command (NORAD)core — airspace/governmental partnerNORAD uses WindBorne Live to track WindBorne's balloon constellation as part of airspace awareness and safety coordination; WindBorne provides comprehensive, real-time balloon tracking via its proprietary software to government partners including NORAD.
Scale indicators16 records
Recent moves7 records
Expansion highlights5 records
WindBorne Systems competitors and assessment
Company assessmentDirect peers
- Tomorrow.io: AI weather forecasting platform built on a proprietary global sensing network (microwave + radar) and sold via API to energy, utilities, aviation, and government. Closest direct analogue to WindBorne: proprietary observation data + AI model + enterprise API GTM serving the same verticals.
- Spire Global: Operates a proprietary nanosatellite constellation that produces GNSS radio occultation and other in-situ atmospheric observations, then sells weather data and forecasts to government and commercial customers. Same 'own the data + own the model' thesis as WindBorne, but with a space-based rather than stratospheric-balloon sensing layer.
- Climavision: Combines a proprietary gap-filling radar network with AI-driven nowcasting and forecasting models sold to broadcasters, utilities, and government. Directly comparable as a 'new observation asset + AI' challenger to legacy NWS / ECMWF data.
Emerging players
- Salient Predictions: Subseasonal-to-seasonal AI weather forecasting startup focused on the 2-week-to-12-month horizon. Partial overlap with WindBorne's U.S. Navy subseasonal work and with energy/agriculture buyers that need longer-horizon signals.
Broad incumbents
- DTN: Long-established weather, agronomy, and commodity-market data and analytics provider for agriculture, energy, and transportation incumbents. Overlaps with WindBorne's utility, energy-trading, and agriculture segments but as a broad incumbent with a much larger existing customer base.
- The Weather Company (IBM): Largest commercial weather business globally, combining media (Weather.com), enterprise data APIs, and AI forecasting on top of public model output. Broad incumbent that WindBorne must displace or partner with to scale enterprise distribution.
- AccuWeather: Incumbent commercial weather forecast and data provider serving media, government, and enterprise; a long-standing established player that WindBorne competes against for government and enterprise weather-forecast contracts.
- StormGeo: Norwegian weather intelligence provider serving shipping, offshore, energy, and aviation with operational forecasting and routing. Direct overlap with WindBorne's energy and aviation/marine use cases (including the GSB-to-buoy marine play), but as a broad incumbent rather than an AI-native challenger.
Regional players
- Meteomatics: Swiss-based high-resolution weather data and forecasting provider selling to European energy, aviation, defense, and government customers. Comparable enterprise weather API and vertical solutions, but with a primarily European footprint.
Others
- Jupiter Intelligence: Climate-risk and physical-climate analytics platform for infrastructure, insurance, and financial services. Adjacent, not directly competitive: serves similar financial / energy / utility buyers with climate risk rather than weather forecasting, but competes for the same budget line.
Market position
Strengths4 records
Weaknesses4 records
Competitive moat5 records
Key risks6 records
Key highlights7 records
Customer concentration
WindBorne Systems social profiles
Digital presenceWindBorne Systems compliance and trust
Trust signalCompliance4 records
WindBorne Systems financial estimates
Financial estimateRevenue estimate
Valuation estimate
WindBorne Systems leadership team
Management profileNumber of profiles
Profiles4 records
WindBorne Systems funding detail
Funding detailFunding overview
Funding rounds7 records
Investors13 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
WindBorne Systems M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about WindBorne Systems
What does WindBorne Systems do?
WindBorne Systems designs, manufactures, and operates a constellation of long-duration Global Sounding Balloons (GSBs) called Atlas to collect in-situ atmospheric data over oceans, polar regions, and remote areas where traditional radiosondes do not reach. It then feeds that proprietary data into its own transformer-based AI forecasting model (WeatherMesh) and multi-model blender (MetaMesh), selling forecast and observation outputs to government agencies, defense customers, energy traders, and utilities through a REST API, Python SDK, and tailored vertical solutions.
Is WindBorne Systems a public or private company?
WindBorne Systems is a private company. It is classified as venture growth investor backed and is currently operating.
When was WindBorne Systems founded?
WindBorne Systems was founded in 2019. It employs 51 to 100 people.
Where is WindBorne Systems based?
WindBorne Systems is headquartered in Redwood Shores, United States, in the North America region.
How does WindBorne Systems make money?
Four revenue lines are on record. Atmospheric Observation Data (Data-as-a-Service) is the primary driver. The others are AI Weather Forecast API (WeatherMesh & MetaMesh), government Contracts and industry-Specific Weather Solutions (e.g., Guardian for Utilities).
Who are WindBorne Systems's main competitors?
Direct peers on record are Tomorrow.io, Spire Global and Climavision. Salient Predictions is listed as an emerging player. Broad incumbents are DTN, The Weather Company (IBM), AccuWeather and StormGeo. Meteomatics is listed as a regional player. Jupiter Intelligence is listed as an others.
Does WindBorne Systems have an API?
Yes. WindBorne offers a public REST API providing programmatic access to its WeatherMesh forecast models (Global and 3km/HighRes), MetaMesh blended forecasts, and Atlas observation data, with both JSON and gridded Zarr response formats. Authentication is via client credentials supplied through environment variables (WB_CLIENT_ID and WB_API_KEY). The API supports point forecasts, gridded forecasts, custom variable queries, sub-second latency, and 1-year backtest data, plus an interactive Insights platform for meteorological-grade visualization. Developer documentation is at api.windbornesystems.com.
What industry is WindBorne Systems in?
WindBorne Systems's product category is AI Weather Intelligence / Atmospheric Data Platform. Its primary akta.pro industry code is EUAMACAJ, Wind Resource Assessment & Energy Yield Analytics (Met Masts, LiDAR, Modeling). Its NAICS code is 221115 and its SIC code is 3812.